A recent robot industry, which is emerged as one of the core industries that are led by the government, has been greatly developed. In particular, with the development of intelligence robot, new technologies are poured out day by day. In particular, an interaction field with human has been remarkably developed. In order to implement the smooth interaction with human, the robot needs to include basic functions of searching and following persons.
Robot services that allow a robot to recognize a person and follow the person, allow a robot to see an external intruder and track the intruder, or the like, may be considered as essential technologies to implement the interaction between a human and a robot.
Various researches for a method for allowing a robot to track a person have been conducted. Recently, an image tracking technology using color information has been mainly used.
Generally, the image tracking technology using color information constructs targeted colors using color models known as being slightly endured to an illumination change such as a normalized red (R)-green (G)-blue (B) space or a hue (H)-saturation (S)-intensity (I) space, or the like.
However, the color models are well adapted to the uniform illumination change of the targeted object, but cannot be adapted to an irregular change in brightness of the targeted object according to an angle change or a sudden change in brightness between illumination and the targeted object, which may cause a failure in tracking a person.
When the person deviates from the viewing angle of the camera, the robot misses a person having been tracked, which may cause the failure in the tracking of the person.